Research is needed to determine the prevalence of and to evaluate employee assistance programs (EAPs) for substance-abusing workers. This information will guide employers to establish the appropriate EAP for their work site. We propose a study to determine work site EAP prevalence and profiles and to evaluate EAP models. We will examine EAP costs and outcomes and compare EAP models across various industries and work sites. Our Methodology has three components: a baseline national survey, case studies, and a follow-up national survey. The national surveys will focus on EAP prevalence, characteristics, use, costs, and outcomes. We will select a national probability sample of work sites from the Dun and Bradstreet sampling frame. The sample selection will be stratified by four work site sizes and six industry groups. We will use our Computer Assisted Telephone Interview (CATI) system to administer the questionnaire to 4,875 work sites with 50 or more employees. After collecting the baseline data, we will select 24 work sites for case study. We will gather cost and outcome information and estimate the cost-effectiveness of EAPs in several areas, including absenteeism, injuries, turnover, and accidents. To bolster information gathered from EAP administrators, we will conduct an employee survey to assess substance abuse and attitudes, and perceptions and use of the work site EAPs. We will also conduct a national follow-up survey (4,875 work sites) to determine changes in prevalence, administration, and use of substance abuse programs. The results of this study will provide insight into the design of EAPs and their impact on work site substance abuse. This composite EAP information will prove useful to both employers and policymakers.